There is an urgent need for a new glucose sensor that is: (1) highly reliable and continuous or near-continuous; (2) capable of automatic function, independent of user initiative, to warn of hypoglycemia; and (3) convenient and acceptable to users. The investigators have an implantable electrochemical glucose sensor that can potentially meet this need. A prototype of this sensor has operated continuously as a central venous implant in dogs for over 100 days without the need for recalibration. This achievement still represents the published world's record for glucose sensor longevity. They are adapting this sensor for long-term use in humans as a tissue implant. The implanted part of the system will include a sensor, electronics and telemetry "chip," antenna, and battery, all of which will be mounted together in a sensor assembly. The glucose sensor signal will be transmitted to an external receiver resembling a beeper that will display the actual glucose value, as well as function as a hypoglycemia alarm. Preliminary studies have shown that tissue structure in the vicinity of the sensor and the wound healing process have major effects on glucose transport to the sensor, and thus affect the sensor signal. There is a need for applied research on glucose transport through tissues and signal interpretation, aspects that must go hand-in-hand. The general aim of this project is to develop a mechanistic understanding of the sensor response to glucose in tissues. The working hypothesis is: "A useful analytical relationship between the signal of sensors implanted in tissue and blood glucose concentration can be found." The investigators have defined a set of rational performance goals based on the relationship between measurement error dynamic lag, and diffusional distances. Studies to be performed include: fabrication of sensors with specified response characteristics; implantation of sensors in well-defined animal models; documentation of the tissue/sensor interface morphology with time; development of methods for signal interpretation; establishment of recalibration methods; development of a comprehensive, sensor- independent model of the tissue sensor response; and refinement of the sensor design as necessary to optimize response. It is expected that these studies will result in a definitive understanding of sensor function as an implant in tissue.